Is preferred candidate polling useful in predicting the likelihood of a candidate winning the primaries?

Question

I've heard repeatedly that Donald Trump is the front-runner for the Republican party presidential primaries when measured by polling for preferred candidate of republican voters, and Wikipedia backs that claim up.

However, I suspect that Donald Trump may not win the primaries, because while he has a large number of people who are strongly favourable to him, there's an even larger number of people who are strongly unfavourable to him.

I suspect that if polling measured what proportion of party voters had an overall positive, neutral, or only mildly negative attitude towards a given candidate, Trump wouldn't fare as well as other candidates.

My question is: Is polling for a party member's number 1 favourite candidate a useful way of predicting how likely a candidate is to win the primary? If not, what other polling is publicly available that can predict how likely a candidate is of winning the primary?

(Note: I'm not trying to ask who will win the Republican primary - I think that as of October 2015, the field is far too open to make predictions)

Answer 1

The primaries are a conventional single-vote election. The voters in the primary can not express disagreement with a particular candidate except by voting for a different one. That means the current prognosis system - asking potential voters who they prefer the most out of all available candidates - is a pretty good indicator.

But remember that in many (not all) states, the Republican party uses the same winner-takes-all system which the actual presidential election uses (list of states by republican primary delegation system). That means when you look at nation-wide polling results, you don't get a prognosis which is as reliable as when you look at the results of all individual states, consider how their delegation systems work, and build a prognosis based on the delegates each state is expected to send.

Also keep in mind that primaries (especially closed primaries where only party supporters can vote) have a tactical component. The parties don't just try to find their favorite candidate, but also the candidate who has the best chance to beat the candidate of the other party. To win the crucial swing-states you need a candidate who appeals to the swing-voters. A highly polarizing candidate who is a hardliner for the party doctrine might be popular within the party but not that popular with independent voters. This might prevent tactically-minded party members to vote for that person during the primaries. When you ask someone "who would you prefer as a president?" and "who will you vote for in the primaries?", the answers might be different.

So when you read any opinion polls in the media, carefully look at what exactly the question was.